HORTSCIENCE 39(3):541–544. 2004. experiment was a randomized complete block with five treatments, six blocks, and five- treatment replicates. Each block comprised Influence of Girdling and Gibberellic a partial vineyard row. Treatments were 1) control; 2) girdle (cane girdle year 1; cordon

Acid on Yield Components, girdle years 2–3); 3) 15 ppm GA3 at bloom; 4) treatment 3 + 40 ppm GA3 14 d later; and 5) treatment 4 + 40 ppm GA 2 weeks later. Composition, and Vestigial Seed 3 Girdling was included as a nonchemical vehicle for possibly increasing berry size and reducing Formation of ‘Sovereign Coronation’ vestigial seeds. Complete girdles were made on the canes and cordons at bloom by girdling Table pliers. GA was applied in the form of Activol 3 as a dilute spray to runoff using a backpack Andrew G. Reynolds and Christiane de Savigny sprayer (10L; Solo, Newport News, Va.). The Cool Climate Oenology and Institute, Brock University, St. same were used throughout the trial. Catharines, ON L2S 3A1 Canada Harvest occurred each year at the discretion of the grower. Data were collected at harvest Additional index words. Seedlessness, growth regulators, Vitis spp., GA (7, 10, and 25 Aug. 1998, 1999, and 2000, Abstract. Vestigial seeds of ‘Sovereign Coronation’ table grapes frequently form partial respectively) for yield components (yield per seedcoats that are perceptible during consumption. This problem was addressed through vine, clusters per vine, cluster weight, berries cane/cordon girdling and gibberellic acid (GA ) sprays. ‘Sovereign Coronation’ vines were per cluster, and berry weight). Cluster weight 3 was calculated from clusters per vine and subjected to one of five treatments [untreated control; cane/cordon girdled; 15 ppm GA3 at bloom (GA1); GA1 + 40 ppm GA 14 days later (GA2); GA2 + 40 ppm GA 14 days yield data. A 100-berry sample was collected 3 3 from each vine and mean berry weight was later]. GA3 had no effect on yield or clusters per vine, but postbloom GA3 treatments increased cluster and berry weights and reduced berries per cluster. Fruit maturity was determined from these samples. Berries per not consistently affected by the treatments, although slight increases in °Brix and pH and cluster was calculated from cluster weight decreases in titratable acidity (TA) were associated with postbloom GA treatments. Use and berry weight data. An additional 10-berry 3 sample was collected annually from each vine of postbloom GA3 applications reduced the number and weight of vestigial seeds with developed seedcoats, and reduced the number and weight of undeveloped seeds as well in before harvest to determine vestigial seed 2 of 3 years. Girdling increased cluster and berry weights, decreased °Brix and TA, and (seed trace) number and weight. Seed traces increased pH. Transpiration rate of on girdled vines was also higher than control were divided into two categories: those with or without partial seedcoat development. Seeds vines on one sampling date. Data suggest that use of bloom and postbloom GA3 applica- tions to ‘Sovereign Coronation’ may reduce the formation of perceptible vestigial seeds with developed seedcoats were clearly brown and thus improve the marketability of this cultivar. in color, whereas those without seedcoats were green and easily discernible strictly on the basis of color from those with seedcoats. Fruit Since its release in 1977 (Denby, 1977), in BC with the GA3 product Activol (Zeneca composition [°Brix, titratable acidity (TA), pH] the stenospermocarpic seedless cultivar Agrochemicals, Fernhurst, Haslemere, Sur- was measured on 100-berry samples using an Sovereign Coronation (Patricia x Himrod, rey, U.K.) yielded similarly promising results Abbé refractometer (AO Corp., Buffalo, N.Y.), Summerland, 1966) has become popular in (Vielvoye, 1992). manual titration to pH 8.30 endpoint, and pH both Ontario and British Columbia (BC). There is an expanding market for ‘Sov- meter (Accumet 25; Fisher Scientific, Nepean, However, adverse environmental conditions ereign Coronation’ and other more recently- Ont.), respectively. Weight of cane can frequently lead to partial seedcoat forma- introduced winter-hardy seedless table grape (vine size) was measured on each vine with a tion on the vestigial seeds. Consumers find varieties (Reynolds et al., 1997a, 1997b). manual dairy scale immediately after this feature disconcerting when they are led to However, there is concern that adverse envi- each season. believe they are purchasing seedless fruit. ronmental conditions (e.g., temperatures <10 Roper and Williams (1989) suggested that Prebloom gibberellic acid (GA3) applica- °C, precipitation, or both) at berry set may girdling of ‘Thompson Seedless’ vines reduced tions to seeded Vitis labruscana Bailey hybrids occasionally lead to full or partial seedcoat transpiration. To confirm this response under results in a high percentage of seedless berries formation on vestigial seeds. The purpose of the much cooler and more humid conditions of (Fukunaga and Kurooka, 1987; Kimura et al., this experiment was to ascertain the efficacy this trial, transpiration (µmol·cm–2·s–1 of H O) 1996; Tonietto et al., 1983; Ueda and Naito, 2 of cane/cordon girdling and GA3 applications was measured hourly on 10 recently-expanded, 1985). The phenomenon is caused by GA3- on seed trace formation and berry size of exposed leaves per treatment replication (con- induced death of the ovule before bloom, after ‘Sovereign Coronation’ table grapes, with the trol and cordon girdled only) on three dates which subsequent GA3 applications enhance ultimate goal of extending the registration of during Summer 2000 using a steady-state berry size through stimulation of cell division Activol in Canada to include grapes. porometer (LI-1600; LI-COR, Lincoln, Nebr.). (Motomura and Ito, 1972). GA3 application Leaves were tagged before measurement to to seeded V. labruscana table grapes such as Materials and Methods ensure that the same leaves were used at each ‘Delaware’ has become a standard cultural measurement time on each date. practice in Japan, and has been mentioned as The trial was initiated in June 1998 in a All data were subjected to analysis of vari- a possible tool for V. vinifera in New Zealand 5-year-old block of ‘Sovereign Coronation’ ance using the SAS statistical package (SAS and Australia (Dry and Smart 1986). In Florida, at Tregunno Vineyards, Queenston, Ont. Vines Institute, Cary, N.C.). The General Linear Mod- the seedless table grape ‘Orlando Seedless’ was were grafted to SO4, spaced 1.5 m × 3.0 m els Procedure was used. Means were separated reported to respond favorably to postbloom (vine × row), and trained to a modified Geneva using Duncan’s multiple range test. GA3 sprays in terms of reduced seed trace Double Curtain system (Shaulis et al., 1966) weight and increased berry size (Halbrooks which included a 2.2-m high bilateral cordon Results and Discussion and Crovetti 1989; Halbrooks and Mortensen containing five-node spurs (upper ), plus 1987; 1988). Trials on ‘Sovereign Coronation’ two additional six-node canes tied horizontally Yield components and vine size. Yield was to fruiting wires at each of 1.0-m and 1.5-m unaffected by girdling or GA3 compared to the Received for publication 2 Jan. 2003. Accepted for heights (lower canopy). The node density at control throughout the trial (Table 1). There publication 18 June 2003. pruning was 15 nodes/m for both canopies. The were also no differences amongst treatments

HORTSCIENCE VOL. 39(3) JUNE 2004 541 GROWTH REGULATORS

Table 1. Influence of girdling and gibberellic acid on yield components and vine size of ‘Sovereign Coronation’ table grapes, Queenston, Ont., 1998–2000. Yield Cluster wt Berry wt Cane prunings wt (kg per vine) Clusters/vine (g) Berries/cluster (g) (kg/vine) Treatment 1998 1999 2000 1998 1999 2000 1998 1999 2000 1998 1999 2000 1998 1999 2000 1998 1999 2000 Untreated control 7.2 7.4 8.4 ab 60 139 113 a 120.2 c 52.6 b 74.1 b 45 28 a 31 2.61 e 1.89 e 2.41 d 1.01 0.71 bc 0.43 Girdled at bloom 8.1 7.3 7.9 b 63 124 96b 130.4 bc 59.0 a 82.5 a 47 28 a 31 2.75 c 2.12 c 2.63 b 0.91 0.66 c 0.39

10 ppm GA3 at bloom (GA1) 7.3 6.9 8.4 ab 60 135 111 ab 121.8 c 52.6 b 76.9 b 47 25 b 30 2.63 d 2.09 d 2.53 c 1.08 0.78 a–c 0.44

GA1 + 40 ppm GA3 14 d postbloom (GA2) 8.2 7.3 8.5 ab 60 126 100 ab 140.2 ab 59.0 a 84.6 a 47 26 ab 30 3.07 b 2.30 b 2.80 a 1.05 0.87 ab 0.43

GA2 + 40 ppm GA3 28 d postbloom 8.1 7.2 9.2 a 57 119 109 ab 143.7 a 60.7 a 85.7 a 47 25 b 31 3.1 0 a 2.44 a 2.79 a 1.1 0.93a 0.48 Significance NS NS * NS NS * *** ** *** NS * NS *** *** *** NS * NS NS,*,**,***Nonsignificant or significant at P ≤ 0.05, 0.01, and 0.001, respectively. Means within columns followed by different letters are significant at P ≤ 0.05, Duncan’s multiple range test. for cluster number in 1998 and 1999, or for Table 2. Influence of girdling and gibberellic acid on fruit composition of ‘Sovereign Coronation’ table berry number and vine size in 1998 and 2000 grapes, Queenston, Ont., 1998–2000. (Table 1). Berry and cluster weight were the oBrix Titratable acidity (g/L) pH two yield components most affected by GA3 Treatment 1998 1999 2000 1998 1999 2000 1998 1999 2000 application and girdling. Berry weights were Untreated control 13.3 a 15.1 c 14.9 b 17.3 a 15.1 a 18.0 b 2.85 c 2.89 c 2.77 b increased over the control by all treatments, Cane girdled at bloom 12.9 c 14.7 d 15.1 a 16.8 c 15.1 a 17.7 bc 2.87 a 2.88 c 2.79 a but in particular by girdling and by either two 10 ppm GA3 at bloom (GA1) 12.7 d 15.3 b 14.8 b 17.3 a 15.0 a 17.8 b 2.85 c 2.91 b 2.79 a GA1 + 40 ppm GA or three GA3 applications (Table 1). A sub- 3 stantial increase in berry and cluster weight 14 d postbloom (GA2) 13.3 a 15.4 a 14.5 c 16.4 d 14.4 b 17.4 c 2.87 a 2.89 c 2.78 b GA2 + 40 ppm GA occurred in treatments GA2 and GA3 (two 3 28 d postbloom 13.1 b 15.1 c 14.4 c 17.1 b 14.1 b 18.4 a 2.86 b 2.93 a 2.79 a and three GA3 applications), as well as the cordon-girdled vines in 1999 (Table 1). There Significance *** *** *** *** *** *** *** *** ** **,*** was also a slight decrease in berry number in Significant at P ≤ 0.01 or 0.001, respectively. Means within columns followed by different letters are significant at P ≤ 0.05, Duncan’s multiple range test. GA3-treated vines (treatments GA1 and GA3) in 1999 (Table 1). Table 3. Transpiration rates (µmol·cm2·s-1 of H O) of The girdling and GA3-induced increases effect on TA in 1998–2000 or on pH in 1998. 2 in cluster and berry weights of seedless V. However, at least one of the two postbloom ‘Sovereign Coronation’ grapevines, Queenston, Ont., 2000. vinifera table grapes are well documented GA3 treatments reduced TA and increased (Weaver and McCune, 1959, 1961; Winkler pH each season (Table 2). Girdling reduced Treatment 19 July 26 July 2 Aug. et al., 1974). Moreover, single applications of °Brix in 2 of 3 seasons, decreased TA slightly Untreated control 15.90 6.48 8.50 Cordon girdled 15.73 7.63 9.96 GA3 at or around bloom have led to increased in 1999, and increased pH in 1998 (Table 2). Significance NS NS *** berry and cluster weight as a result of berry Applications of GA3 have been reported to thinning, but the berry weight increase is less advance fruit maturity (Christodoulou et al., NS,***Nonsignificant or significant at P ≤ 0.001, than that induced by later applications (Christo- 1968), delay maturity slightly (Christodoulou respectively. doulou et al., 1968). Use of GA3 on seedless et al., 1968), or have had no effect (Weaver and V. labruscana hybrids has also proven to have McCune, 1961; Weaver et al., 1962), depending similar effects (Shaulis, 1959). Vielvoye (1992) upon cultivar and season. Previous work with and weight (Fig. 1 E and F). Girdling had no reported that girdling and one or two 40-ppm ‘Sovereign Coronation’ in BC using similar effect on vestigial seed formation throughout the trial (Fig. 1). postbloom GA3 applications increased cluster application timings suggested an enhancement It is also noteworthy that vestigial seed and berry weights of ‘Sovereign Coronation’ in fruit maturity with both GA3 and girdling in a similar experiment in British Columbia, (Vielvoye, 1992). number and weight varied considerably during but a single 10-ppm application at bloom had Vestigial seed formation. As with previous the 3 years of the trial. In the dry 1998 and little impact. experience with stenospermocarpic varieties 1999 seasons, mean mature seed numbers in A slight increase was also measured in (Halbrooks and Crovetti, 1989; Halbrooks and control treatments were 1.88 and 1.10, respec- tively, but only 0.12 in the much wetter 2000 vine size in 1999 in vines treated with GA3 Mortensen, 1987; 1988; Vielvoye, 1992), our season (Fig. 1 A, C, and E). Moreover, mean the previous season (Table 1). Vine size de- data suggest that GA3 reduced vestigial seed weights of mature seeds in the control treat- creased from 1.03 kg/vine across all treatments formation (Fig. 1). Use of postbloom GA3 in 1998 to 0.43 kg/vine in 2000, suggesting applications tended to reduce the number and ments were 24.4, 15.9, and 0.9 mg for 1998, that the vines may have been overcropped in weight of seeds with developed (brown, hard) 1999, and 2000, respectively, suggesting that the final year of the trial. The reason for the seedcoats in 1998 (both GA2 and GA3 treat- environmental conditions, in this case lower reduction in vine size is not known; however, ments; Fig 1 A and B) and 1999 (GA2 only; than average rainfall (Ontario Grape Growers’ a 28% higher-than-average rainfall in 2000 Fig. 1 C and D) and concomitantly increased Marketing Board, 2001), might have played a (637 mm, compared to an average of 498 mm) the number of undeveloped (green, soft) seeds part in their formation. It is also possible that (Ontario Grape Growers’ Marketing Board, in 1998 (Fig. 1A). In 1999, the treatment involv- the high crop loads (yield : vine size ratios) in 2000 (19.8) compared to those in 1998 and 2001) with attendant soil waterlogging, may ing three GA3 sprays also reduced the number have been a contributing factor in limiting and weight of green seeds (Fig. 1 C and D). 1999 (7.6 and 9.1, respectively) may have reduced vestigial seed formation as well. vegetative growth. However, use of a third GA3 spray in 1999 did Fruit composition. Fruit composition data not effectively suppress formation of mature Water relations. Two of three sampling days produced no differences between un- suggested that the GA3 treatments had either vestigial seeds when compared to a two-spray no effect or led to slight delays in fruit maturity program (Fig. 1 C and D). In 2000, vestigial treated control and girdled vines with respect in terms of soluble solids (°Brix), with the seed number and weight were substantially to transpiration rates (Table 3). However, exception of the GA1 and GA2 treatments lower than the previous 2 years, and there the third sampling date suggested a slightly higher transpiration rate in girdled vines. in 1999 (Table 2). The single bloom GA3 ap- were no differences relative to the control with plication reduced °Brix in 1998, and had no respect to both brown and green seed number This is in contrast to prior work (Roper and

542 HORTSCIENCE VOL. 39(3) JUNE 2004 Williams, 1989) and others who have reported one bloom spray plus two or three postbloom had no effect on vestigial seed formation in decreases in stomatal conductance (hence applications of GA3 may be used to reduce the this trial, but it increased cluster weight and transpiration) of leaves on girdled vines. The number and weight of vestigial seeds in ‘Sov- berry weight in all three seasons. These cultural reason for higher stomatal conductance in ereign Coronation’ table grapes. Moreover, use practices should provide table grape growers girdled ‘Sovereign Coronation’ in this study of postbloom GA3 sprays will also increase in Canada and the northeast U.S, some com- is unknown. the probability of higher berry weight and petitive advantage in a market dominated by The results of this experiment suggest that cluster weight than untreated vines. Girdling V. vinifera cultivars.

Fig. 1. Influence of girdling and gibberellic acid on vestigial seed formation in berries of ‘Sovereign Coronation’ table grapes, Queenston, Ont., 1998–2000.

Treatment legend: GA1 = 10 ppm GA3 at bloom; GA2= GA1 + 40 ppm GA3 14 d. postbloom; GA3 = GA2 + 40 ppm GA3 28 d postbloom. Means followed by different letters are significantly different at P ≤ 0.05, Duncan’s multiple range test. Significant P values for 1998–2000 were, respectively: 0.0001, 0.005, and 0.021 (green seed number); 0.0001, 0.026, and 0.050 (green seed weight); 0.0001, 0.050, and 0.700 (mature seed number); 0.0001, 0.050, and 0.636 (mature seed weight).

HORTSCIENCE VOL. 39(3) JUNE 2004 543 GROWTH REGULATORS

Literature Cited Effects of gibberellic acid and streptomycin on Response of Concord grapes to light, exposure, Christodoulou, A.J., R.J. Weaver, and R.M. Pool. pollen germination and ovule and seed develop- and Geneva Double Curtain training. Proc. Amer. 1968. Relation of gibberellin treatment to fruit- ment in Muscat Bailey A. Amer. J. Enol. Viticult. Soc. Hort. Sci. 89:268–280. set, berry development, and cluster compactness 47:152–156. Tonietto, J., A. Miele, and P. Silveira Jr. 1983. O in Vitis vinifera grapes. Proc. Amer. Soc. Hort. Motomura, Y. and H. Ito. 1972. Exogenous acido giberelico no desenvolvimento de bagas Sci. 92:301–310. gibberellin as responsible for the seedless berry sem sementes da uva ‘Isabel’. Pesq. Agropec. Denby, L.G. 1977. ‘Sovereign Coronation’ grape. development of grapes. II. Role and effect of the Bras. 18:381–386. HortScience 12:512. prebloom gibberellin application as concerned Ueda, H. and R. Naito. 1985. Relation between the Dry, P. and R.E. Smart. 1986. How to get rid of the with the flowering, seedlessness, and seedless type of flowering and the seedlessness ratio in the seed in seeded table grapes. Austral. Grapegrower berry development of Delaware and Campbell GA-treated clusters of ‘Muscat Bailey A’ grape. & Winemaker (Oct.):22. Early grapes. Tohoku J. Agr. Res. 23:15–23. J. Jpn. Soc. Hort. Sci. 54:192–200. Fukunaga, S. and H. Kurooka. 1987. Studies on Ontario Grape Growers’ Marketing Board. 2001. An- Vielvoye, J. 1992. Winemakers anticipate “vintage seedlessness of ‘Kyoho’ grapes induced by nual report and financial statements for the year of the century”. B.C. Grape Wine Ind. Nwsltr. st gibberellin in combination with streptomycin. ended January 31 , 2001. Ont. Grape Growers’ 2(6):1. Bul. Univ. Osaka Prefec. Ser. B 40:1–10. Mktg. Board, Vineland Station, Ont. Weaver, R.J., A.N. Kasimatis, and S.B. McCune. Halbrooks, M.C. and A.J. Crovetti. 1989. Gibberellic Reynolds, A.G., M.J. Bouthillier, D.A. Wardle, and 1962. Studies with gibberellin on wine grapes acid increases berry size, reduces seed traces in L.G. Denby. 1997a. ‘Skookum Seedless’ grape. to decrease bunch rot. Amer. J. Enol. Viticult. ‘Orlando Seedless’ grape. Proc. Plant Growth HortScience 32:743–744. 13:78–82. Reg. Soc. Amer. 16:34–39. Reynolds, A.G., M.J. Bouthillier, D.A. Wardle, and Weaver, R.J. and S.B. McCune. 1959. Effect of Halbrooks, M.C. and J.A. Mortensen. 1987. Influ- L.G. Denby. 1997b. ‘Sooke Seedless’ grape. gibberellin on seedless Vitis vinifera. Hilgardia ence of gibberellic acid and various management HortScience 32:745–746. 29:247–275. Roper, T.R. and L.E. Williams. 1989. Net CO Weaver, R.J. and S.B. McCune. 1961. Effect of practices on berry, seed, and cluster development 2 in Orlando Seedless grape. Proc. Fla. State Hort. assimilation and partitioning of gibberellin on vine behavior and crop production Soc. 100:312–315. grapevine leaves in response to trunk girdling in seeded and seedless Vitis vinifera. Hilgardia Halbrooks, M.C. and J.A. Mortensen. 1988. Effects and gibberellic acid application. Plant Physiol. 30:425–443. of gibberellic acid on berry and seed develop- 89:1136–1140. Winkler, A.J., J.A. Cook, W.M. Kliewer, and L.A. ment in Orlando Seedless grape. HortScience Shaulis, N.J. 1959. Gibberellin trials for New York Lider. 1974. General viticulture. Univ. Calif. 23:409. grapes. N.Y. Farm Res. 25:11. Press, Berkeley. Kimura, P.H., G. Okamoto, and K. Hirano. 1996. Shaulis, N.J., H. Amberg, and D. Crowe. 1966.

544 HORTSCIENCE VOL. 39(3) JUNE 2004